Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Population biology of the weasel Mustela nivalis on British game estates

Population biology of the weasel Mustela nivalis on British game estates King, C. M. 1980. Population biology of the weasel Mustela nivalis on British game estates. - Hotarct. Ecol. 3: 160-168. The mean year-class ratio in 6 samples of weasels (total n = 477) ranged from 59% to 84% young, and mean age from 0.79 to 1.16 yr. There was no difference in age structure between samples from 5 game estates and 1 reserve. Mean annual mortality of weasels of both sexes and all ages was 75-90%. In 12 pregnant weasels the mean number of embryos, observed from April to June, was 5.6. There was a defmite anoestrous season in winter. Records of the number of weasels killed on game estates show that weasels and stoats reacted differently to the first outbreak of myxomatosis in England in 1954; and that weasel populations are capable of enormous sudden increases (usually associated with "mouse years"). These increases are probably due largely to summer litters produced by precocious juveniles, and also to some adults breeding a second time (both possibilities confirmed from ovarian histology on this material). Traditional gamekeeping practice may influence the seasonal pattern of mortality of weasels (highest in spring) but its effects are probably short-lived. C. M. King, 3 Waerenga Road, Eastbourne, New Zealand. 1. Introduction Published studies of the population biology of Mustela nivalis are few, mainly because the two essential requisites, an accurate method of age-determination and unbiased samples, are difficult to supply. The age of weasels is difficult to determine (King 1980b), and in Britain, though large samples may be collected from co-operative gamekeepers, they are usually biased in several different ways. The literature does not contain a single analysis of the dynamics of a wild population of Musiela nivalis, not even a first approximation, although the commercially valuable mustelids have received closer attention, especially in the USSR (see, for example, many of the papers on sable, marten and ermine translated in King, 1975c, 1980a). In Britain the population biology of weasels is of considerable interest to the game industry, though the lack of any population data hitherto has prevented rational assessment of the traditional methods Throughout this paper, the name "weasel" refers only to Mustela nivalis, and "stoat" to M. erminea. Accepted 3 September 1979 © HOLARCTIC ECOLOGY 0105-9327/80/030160-09 S 02.50/0 usually employed for "vermin control". The present study sketches a first outline of this important missing information. Published estimates of age ratios for Mustela nivalis are difficult to compare with present results because not all authors defined their terms and methods in sufficient detail. Since first-year weasels* comprise about threequarters of the population, the total number at any time will depend largely on the production and survival of young, and the proportion of young in the sample will depend on the season of collecting and the definition of the youngest age class(es). Lockie (1966), Hansson (1968), Fog (1969), Barbu (1968) and Stubbe (1969) all used different definitions of age. Comparing them all shows only that all authors except Barbu and Hansson found the young animals to be in the majority. The distribution and abundance of weasels is clearly related to the local population of small rodents (Rubina I960, Parovshchikov 1963, Heptner et al. 1967, Eriinge 1974, King 1975a), though there are few direct data on the influence of nutrition on productivity (Tapper 1979), and none on its effect on survival. Reproductive cycles and development have been described by HJII HOLARCTIC ECOLOGY 3:3 (1980) median birth date assumed to be 1 June, although the season of births, beginning in April, is variable in length according to environmental conditions. The few old animals were all given an arbitrary age of 30 months. The weasel is a short-lived animal, but the age structures are expressed in year-class ratios, for practical reasons. The method of age determination has to work in chronological categories, because young weasels vary 2. Material and methods greatly in the age at which they reach reproductive Five of the samples (455 weasels) were collected by maturity, according to when in the season they were gamekeepers on estates in England and Scotland (num- horn. The year classes can of course be subdivided by bers 1, 2, 5, 6 and 7 on the map in King 1977). All were month, but if this is done, as in Tab. 4, the season at killed in Fenn traps, described by King and Edgar which the sample was collected determines the relative (1977), and were from populations regularly cropped prominence of each monthly or quarterly class. A samby trapping. A small sample of 22 weasels from ple collected in summer (June-August) contains many Wytham Woods, Berkshire, a reserve with no history of juveniles and few sub-adults, and vice versa one colweasel-trapping before 1968, is considered separately: lected in spring, though the same cohort of young, at these animals were either caught in box traps or shot. different stages in their development, is represented in Weasels which had been held in captivity or picked up both. By comparing samples in year-classes, differences due to season of sampling are reduced. Further, as dead were excluded. Most carcases were stored frozen. Autopsy proce- pointed out by Caughley (1977) and illustrated pardures included recording of measurements, food habits ticularly well in weasels, differences between age classes and Skrjabingylus nasicola (King 1977), and counting of spanning less than a year will relate more to season than embryos, and, for the first 171 specimens (samples 4, 5 to chronological age, so classes defined in years are not and 6, plus the first 55 of sample 1) also fatness, moult arbitrary; they are the only natural units of age. pattern, weight of gonads and histology of ovaries. PlaYoung weasels cannot be trapped before four weeks cental scars were not visible (Heidt 1970). Skulls were of age, when they first leave the nest, and are unlikely to cleaned by dermestid beetles. Demographic procedures enter traps before eight weeks, when they can kill for followed, where possible, the advice of Caughley themselves (East and Lockie 1964), Hence, all figures (1977). derived from trapped samples refer to mortality from at As on most game estates, the keepers concentrated least two months, not from birth. The samples are taken on trapping predators in spring, and hence the monthly to represent the standing age distribution in the living distribution of the samples is very uneven: 46% of the population from independence (at about three months) weasels examined were caught during March, April and onwards. May. This sampling programme was not ideal for The problems of population biology can be appopulation analysis, but the resulting material can at proached at several levels of detail, depending on the least be regarded as typical of the annual weasel harvest precision of the answer required and the difficulty of on many British game estates. extracting it (Caughley 1977). In the first instance, the All the game estates were mixed arable farmland at simplest approach is to treat all individuals of a given less than 500 m altitude: the largest sample came from sex or age group as if they were identical and to assume North Farm, Sussex, the study area described by Potts that the population statistics are representative and apand Vickerman (1974). Wytham is a largely deciduous ply as an average over several years. This approach is woodland described in detail by Elton (1966). There the most appropriate one for a pilot study, and is the were no estimates of weasel density on any of the game one used here. estates, but the resident population on a small part of Wytham was observed by livetrapping during 1968-70 (King 1975a). 3. Results The method of age-determination used is based on 3.1. Sex ratio the progression of changes in the shape of the skull established from specimens of known age (King 1977, The great excess of males in the samples (Tab. 1) was 1980b) Skulls were grouped first according to the analysed previously by King (1975b), who considered it month killed, and then into year-classes, divided at 1 more likely to be a sampling error arising from the difJune. Weasels in their first year are called "Young", ferences in the biology of males and females than a and those in their second year "Adults". Adults past measure of the real sex ratio of the population. Howtwo years of age ("Old") may be distinguished by wear ever, there were no seasonal variations in the relative of the carnassial teeth, which is not significant before number of males and females caught (Tab. 2); there was that age; but as this distinction is not precise it is used in no significant difference between males and females in only Tabs 1, 3 and 4. Mean age was estimated from a age structure as estimated from these trapped samples HOLARCTIC ECOLOGY 3;3 (t980) (1939), Deanesley (1944), and Heidt (1970); and variations in fur, bounty or vermin records by Parovshchikov (1963), Jefferies and Pendlebury (1968), and Hewson (1972). Tab. 1. Age distributions of weasels from 5 game estates and 1 reserve. Sussex Oxon. Game estates Northumb, Wigtown. Aberdeen Estates Combined Reserve Wytham Males Young Adult Old Total 164 39 39 30 Difference in age structure between estates (males): X = 2.75, 4 df (adult and old year classes combined), NS. ^ Mean age 0.86 yr Females Young Adult Old Total 66 13 5 84 «7 23 6 Difference in age structure between estates (females): X' = 4.82, 4 df, NS. Mean age 0.93 yr Total 248 42 54 33 78 455 22 Difference in age structure between sexes on the estates: X = 1.05, 1 df, NS. ^ Difference in age structure between estates and reserve (sexes pooled) X * 85.9 0.79 83.3 % males Mean age (yr) %young Ha/kill trap Collected Sample number (King 1977) 5,3 1968-72 1 ? 1968-70 2 66,7 8.1 1968-70 5 9.7 1969-70 6 74.5 0.88 78.5 1968-72 1970-72 1968-70 4 (Tab. 1); and sex ratio did not change with age (King 1975b). For the purposes of this paper I assume that both males and females have been sampled randomly, but at different levels, and so the results for the two sexes are presented separately below. 3.2. Age structure In every sample most of the weasels examined were in the first yearclass (Tab. 1). The actual ages of the old (2 years+) animals is unknown, but the age structures suggest that very few of them were much past their third year. The differences in age structure within the game estates, and between the game estates together and the reserve, were not significant (Tab. 1). The individual weasels examined lived on average about one year or less (the range of mean ages observed was 0.79-1,16 yr, mean 0.88 yr, in different samples, and 0.86 in males, 0.93 in females). However, the mean for the total population (including the pre-independence age classes, which was not sampled) was no doubt much less. The mean expectation of life at birth cannot be estimated if mortality before independence is un162 known. Weasels in captivity may live up to ten years (Frank, pers. comm.). The youngest weasels collected by gamekeepers were all over fifty days old, as all had completed their permanent dentition (East and Lockie 1964). This is probably not because younger weasels cannot be caught, but because the gamekeepers on the estates studied paid less attention to trapping in summer. In Wytham, one young male with deciduous canines still in place was caught in a live-trap in September 1969. A change in the proportions of the year-classes with time may indicate differential vulnerability to trapping (for example, in sable, many more young are caught at the beginning of the season than at the end: Mel'nikov 1975. In the present material, the proportion of young, in all samples combined, fluctuated irregularly through the year, with no obvious decline with time since the season of births (Tab. 2). The mean annual mortality rates of the two year classes were similar, in both males and females (Tab. 3). If regular trapping were having any effect upon the target population, e.g. by stimulating breeding or immigration, or reducing the mean lifespan, a higher HOLARCnC ECOLOGY 3:3 (1980) Tab, 2. Proportion of young and adult weasels and sex ratios through the year (Game estates combined). Age Tab. 4. Seasonal pattern of mortality in weasels on game estates. Capture date Males First year Jun-Aug (Summer) Sep-Nov (Autumn) Dec-Feb (Winter) Mar-May (Spring) Second year Jun-Aug Sep-Nov Dec-Feb Mar-May Older (all seasons) Females First year Jun-Aug Sep-Nov Dec-Feb Mar-May f. 339 293 245 193 Young Jun Jul Aug Sep Ott Nov Dec Jan Feb Mar Apr May 15 32 18 II 18 32 34 24 18 82 54 36 Total %Young 29 43 23 U 22 36 36 32 21 100 69 45 Sex (from King 1975b) n CfCf Total %cfcf U 1.00 0.86 0.72 0.57 d, 0.14 0.14 0.15 0.37 q. 0.14 0.16 0.21 0.65 51.7 74.4 78.3 100 81,8 88,9 94.4 75,0 85.7 82.0 78,3 80,0 76.5 58,8 70.0 100 77,8 77.2 73.7 51.5 80,0 78,0 74,7 73.7 0.20 0,05 0. 15 0.02 0. 13 0.02 0. 11 0,09 0.02 0,82 trapping pressure (expressed as ha/trap) should be Jissociated with a lower mean age. The trapping pressure was estimated for only three of the game estates, and, within the range observed, was not obviously related to variations between estates in mean age (Tab. 1). The mean age in the Wytham Reserve was 1.02 yr (n = 22), which was within the range of mean ages observed on the game estates. The significance of the differences between these figures was not tested because the actual ages were not known: the range of birth-dates (4—5 months) is about the same as the range of differences observed. 3.3. Seasonal (three-monthly) variations in mortality rale (qj 0.15 O.Il 0.22 0.58 0.32 0 0.24 0.62 Second year Jun-Aug Sep-Nov Dec-Feb Mar-May Older (all seasons) mathematically related, but q^ is the only one plotted, as it is the statistic least affected by errors of age-determination and sampling bias (Caughley 1977), 3.4. Natural mortality The mortality pattern of the weasels on the game estates showed a well-defined peak in the spring of each year (Fig. I). However, although this suggests that the activities of gamekeepers may affect the seasonal distribution of mortality, it does not imply that they necessarily affect the annual level of mortality. The q, curve for most mammals is U-shaped; the initial drop is absent from Fig, 1 because q» data derived from trapped samples take no account of mortality before independence. All the colums of Tab. 4 are There is little evidence from this study of the causes of natural mortality in weasels. Of eight weasels found in the pellets of tawny owls in Wytham (Southern, pers. comm.), five were female, two male and one undertermined; two were adult, two young and four undetermined. The nematode parasite Skrjabingylus nasicola causes considerable damage to the sktills of weasels, and 77-100% of skulls in the present samples were infested, but there was no evidence that infested weasels were smaller, lighter, leaner or died younger than uninfested ones (King 1977). Tab. 3. General pattern of mean annual mortality rates of male and female weasels on game estates. Age (yr) Frequency fx Males 0.25-1 1-2 339 69 8 116 29 6 Survival I. 1.00 0.20 0.02 1.00 0.25 0.05 Mortality d. 0,80 0.18 0.80 0,90 Mortality rate Survival rate P« 0,20 0.10 2+ Females 0.25-1 1-2 2+ HOLARCTIC ECOt-OGY 3:3 (1980) Tab. 5. Records of weasel pregnancies and lactations. Pregnancies Date 19 Apr 1953 25 Apr 1958 26 Apr 1971 27 Apr 1972 during april 1969 1 May 1969 10 May 1969 12 May 1969 17 May 1962 21 May 1972 during May 1969 26 Jun 1958 Sample area Hampshire' Hampshire Sussex Aberdeen Sussex Northumberland Northumberland Wytham Hampshire Sussex Sussex Hampshire Body wt (g) No, of embryos 0-2 ADULTS Mean Lactations Sample area Sussex Dates Fig. 1. Seasonal variation in mortality rate, q^, in male (•) and female (o) weasels on game estates. Sm: summer. A: autumn, W: winter, Sp: spring. 3.5. Reproduction 3.5.1. Litter size The mean number of embryos was 5.6 (Tab. 5). As found by Deanesly (1944) the corpora lutea in weasels persist for some lime after parturition (in one pair of ovaries I sectioned, corpora lutea were still present in October, and in another pair, two generations of corpora lutea were found together in July); but the poor quality of the histological material made it impossible to count them. 3.5.2. Breeding season Embryos were observed between April and Jtjne, and lactation between April and July with one (in the south of England) lactating in October (Tab. 5). The end of the season is difficult to determine, because few weasels were collected from mid-summer onwards, and in any case the length of the season probably varies from year to year. Tapper (1976) made a special effort to collect weasels in summer from North Farm, Sussex, and some of those he caught are included in this material. In July and August of 1971 and 1972, 19 females were collected, of which 5 were rotten. As the ovaries were not sectioned, the breeding condition of the remaining 14 could not be fully determined. From macroscopic signs, 2 of the 14 were lactating, and one was in oestrus, but none was visibly pregnant. At Wytham, the longest resident adult female, 9 18, was pregnant when livetrapped on 26 August 1969, but had apparently lost the litter by the time she was next captured three weeks later (King 1975a). The non-fertile season was well marked. In 26 females collected between November and March inclusive, whose ovaries were sectioned, all but one were anoestrous (Tab. 6). The annual cycle of testis weight followed closely the description of Hill (1939). 5 Apr 1971 , 7 May 1971, 1 Jun 1971, 23 Jun 1972, 30 Jun 1971, 19 Jul 1971,24 Jul 1972,14 Oct 1971 Northumberland 9 Jun 1969 Hampshire 2 Jun 1949 1, Hampshire records from Blank (unpubl.). 3.5.3. Fertilily and age Tab. 6 shows the relationship between breeding condition, season and age in the 45 females whose ovaries were sectioned. All three young females caught before the end of the breeding season were fertile, though none had visible embryos. In 16 mature females caught after April, all except one were in breeding cotidition, and the only one caught in July was already fertilised for her second pregnancy; only five of these had visible embryos. Tab. 6. Reproductive condition and age cf 45 female weasels whose ovaries were sectioned. A, Young 9 9 caught before 31 Oct of [heir first year: B, Females of all ages (mature and immature) caught 1 Nov~31 Mar inclusive: C, Females of all ages (all mature) caught after 1 Apr: Fertile (in oestrus or with corpora lutea) 3, Infertile 0. Anoestrous 25 Oesirous 1 (12 Mar) Anoestrous 1 (13 Apr) Oestrous 5 (16 Apr to 9 May) Between ovulation and lactation of first pregnancy of season 9 (28 Apr-9 Jun) Between ovulation and lactation of second pregnancy of season 1 (7 Jul) Infertile 0 Total 16 HOLARCTIC ECOLOGY 3:3 {19B0) Tab. 7. Recent variations in the number of weasels killed per year on three of the game estates studied. Year Sussex Northumberland Wigtownshire (14000 ha) (1200 ha) (4000 ha) none need be expected in these samples, for three reasons. Trapping need have no effect on age ratios if all ages are equally liable to be caught (Caughiey 1974), or if the artificial mortality imposed by gamekeepers is insignificant; and in theoretical populations with similar patterns of age-dependent survival but different rates of increase, the age-ratios can be unrecognisably different (Caughley 1977). Hence, age-ratios cannot be interpreted without a knowledge of r. 4.2. Mortality and survival 3.5.4. Uterus weights Uteri were weighed complete with ovaries attached. The mean weight of 26 anoestrous uteri was 43 mg, range 24—76 mg: of 13 uteri, determined from ovarian sectioning and the appearance of the vulva to be pre- or post-oestrous, 137 mg, range 45-275 mg: of 5 uteri containing embryos, 2.17-6.68 g. 3.5.5. Population Fluctuations The densities of weasels on the five game estates were not known, but they were certainly not stable from year to year. Records of the annual kill of weasels by gamekeepers, which are Ihe crudest form of population estimate, were available from three of the estates for the years preceding the study (Tab. 7). They show that the "harvest" of weasels is unstable, frequently varying between years by a factor of 2 or 3. This instability means that data on mortality and reproduction based on specimens collected over several years provide only a general picture. Tapper (1979) describes the responses of a local weasel population lo changes in its food supply. 4. Discussion 4.1. Age slnidure The pronounced seasonal variation in mortality rate of weasels on game estates is probably a largely artificial consequence of the extra trapping effort traditionally made by gamekeepers in spring. However, weasels also change their behaviour with the onset of the breeding season in spring, and this may increase their "trappability" and also, on untrapped areas, increase the risk of death from factors other than traps. Several resident male weasels in the reserve died after drastic loss of weight at this season (King 1975a). In the combined sample from the game estates, the survival of female weasels from independence to one year old was slightly higher (0.25) than that of males (0.20: Tab. 3), though the difference in the age distributions of the sexes was not statistically significant (Tab. 1). Gamekeepers invariably kill about three times more male weasels than females (King 1975b) which means that more females than males escape the trapping season and later die from natural causes. The small difference in these survival figures perhaps reflects the lower proportion of the female population vulnerable to the risk of being trapped. In contrast, live-trapping studies of undisturbed weasel populations have frequently recorded periods when females were apparently scarce or completely absent (Lockie 1966, Eriinge 1974, King 1975a, Moors pers. comm.). The difference could perhaps be due to the beneficial effects of cropping. Cropped populations often show improved condition and vigour; the effects of traditional weasel trapping, highly selective for males (Tab. 2) would benefit females the most, particularly as in weasels the nonbreeding females are socially subordinate to males and inferior to them in hunting ability (Eriinge 1975), 4.3. Reproduction Stubbe (1969) commented that the high proportion of young in samples of weasels implies high productivity in weasel populations; but immigration stimulated by constant cropping and selective trapping of young can result in an observed proportion of young far exceeding the possible productivity of a population. Hence, the productivity of the population sampled can be estimated only from fecundity data obtained from examining the females. In the present material, only 145 of the total of 535 weasels collected were females, of which only 8 The age structures observed correspond reasonably well with what would be expected in a small mammal capable of rapid reproduction, and with field observations. Few resident weasels hold their territories for more than one year (Eriinge 1974, King 1975a), and none has been known to live in the wild for more than three years (Lockie 1966). Because of the variable occurrence of summer litters, a theoretical age pyramid cannot be constructed for weasels to compare with that calculated for stoats by van Soest and van Bree (1970). There was no difference between the age-structures of the weasels from the estates and the reserve, but HOLARCTIC ECOLOGY 3:3 (1980) were pregnant. Even when sampling is carefully controlled, female weasels are apparently particularly difficult to catch when pregnant (Deanesly 1944, Behnke 1966). The only recorded count of corpora lutea in British weasels is that of Deanesly (1944), who found an average of 7.1 corpora lutea in 32 pregnant weasels examined. The mean number of embryos found in this study {5.6} was comparable to the 6.4 found by Deanesly. The mean number of young born, In 17 litters recorded in Britain, was 6.2 (range 4-8): if a further 22 litters from other parts of the species distribution are included, the mean is 5.0 (King 1975b). The limits of the anoestrous season in British weasels seem to be fairly definite (Deanesly 1944, Tab. 5, 6), in contrast to European nivalis (Stubbe 1969, Heptner et al, 1967) and American rixosa (= nivalis) (Hall 1951). Adult female weasels are known to be capable of producing two litters a year (Heidt 1970), and young females bom early in the season may themselves reproduce later in the summer (Deanesly 1944). Such extended breeding seasons are associated with years when small rodents are very abundant (Hensel 1881) and so also are records of very large litters, which may be up to at least fifteen young (Heptner et al. 1967, Eitzgerald in press). Second litters are possible in "lemming years" even in the short summers of the far north, as Frank (1974) reports that a captive female of the small boreal form M. nivalis rixosa regularly came into post-partum oestrus before the end of lactation, and could rear two litters in 5.5 months. However, monthly trapping records, showing two peaks in capture rate in April and July, do not infer that second litters are normal in Britain (Jefferies and Pendlebury 1968). The first peak is too early to represent dispersing young of the first litter (Tabs 5, 6) and may be more related to spring changes in behaviour and trappability mentioned above. In poor years, many adult weasels may fail to breed altogether (Tapper 1979). 4.4. Population fluclualions Appendix I, Vermin records from Elveden estate, Norfolk (approx 9300 ha). Year Number of animals killed No keepers M. anc warreners M. nivalis erminea rabbits employed 83 92 88 73 88 56 59 168 56 73 75 85 90 34 7 10 11 10 4 3 14 8 20 17 18 I t 17 35 22 42 81 37 16 26 64 22 23 24 715 766 859 925 1013 825 459 419 422 439 596 655 643 434 409 438 569 517 495 855 624 568 735 647 603 645 727 825 257 146 98 80 62 71 40 58 75 170 251 257 166 138 164 225 270 168 249 348 318 448 51563 61571 55986 51777 57240 48602 58561 54169 52606 51163 48870 43416 42294 31706 20348 15765 13218 22744 40376 47742 50009 53793 47645 27381 23535 19528 19839 8601 401 1864 4968 5016 3037 9001 3962 5105 4864 2607 680 849 949 1147 1760 1354 2954 3158 4083 8917 14378 15698 35 Records not available 37 37 33 13 13 14 14 16 26 29 27 27 22 26 25 26 26 19 19 t9 20 19 23 23 22 19 15 11 Keepers on War Service First outbreak of myxomatosis U2 The annual variation in the number of weasels killed on British game estates illustrated in Tab, 7 is comparable with figures published elsewhere: for example, for the ten years 1947-48 to 1956-57 inclusive. Anon (1960) reports the annual kill of weasels on 1336 ha in Hampshire as 82, 106, 185, 96, 84, 110, 46, 64, 53 and 68 respectively. Series of figures showing a similar range of variation can be found in the vermin books of almost any estate, and well back into last century (Middleton 1934). Ocassionally, weasels are capable of sudden quite startling irruptions. The next figure in the above series, for 1957-58, is 348, a five-fold increase over that of the previous year. This is almost certainly due to the consequences of myxomatosis, which were demonstrated in records from all over the country (Jefferies and 166 Pendlebury 1968, Hewson 1972, Appendix 1.) As the rabbits declined, followed by the stoats, the herb layer of the vegetation became luxuriant, and populations of small rodents reached peak levels (Fenner and Ratciiffe 1965, Hewson and Kolb 1973), followed by (he weasels. Over much of the country the main result was a reversal of the previous ratio of stoats to weasels (Craster 1970, Hewson 1972). The increase in weasels was probably due mainly to the improvement in their HOt_ARCTIC ECOLOGY 3:3 (1980) food supplies, though the removal of interspecific competition from stoats may have helped (King and Moors 1979). Appendix 1 suggests that, at least over the last twenty years, the harvest of weasels on this estate has been higher than usual every three or four years. The greatest value of vermin records is that they are an eloquent warning against any facile assumption of a stationary population in short-term or localised samples of weasels. Fur and bounty records in general are often unreliable, because the incentive to set traps changes with the density of the target population; but the figures provided by gamekeepers are probably less subject to this error, since setting traps is done in the course of other work {Hewson and Kolb 1973). It seems safe to assume that weasel populations are actually at least as unstable as these records suggest, and that the rate of increase is usually either positive or negative, but very seldom zero. This means that lifetables more accurate than the preliminary ones presented here can be constructed only by studying marked animals (Caughley 1977). Sudden large population increases in weasels are often associated with years when small rodents are abundant (Swanson and Fryklund 1935, Lokemoen and Higgins 1972, McLean et al. 1974, Goszczynski 1977). This could readily be understood if the effect of extra food is merely to decrease the high mean annual mortality rate; but in weasels, improved survival can be augmented not only by an increase in fecundity, which occurs in other carnivores and may reasonably be expected, but also the capacity (unique among mustelids) of weasels to produce extra litters in good seasons. 4.5. Implicalions for the preservation of game The flexible reproductive strategy of the weasel allows it to avoid surplus production in poor years, and to exploit favourable circumstances very quickly, despite the attentions of gamekeepers. This latent breeding capacity is certainly enough to account for large population peakssuchas that of 1957-58 reported by Anon (1960) and others. In ideal conditions in Britain a single female weasel could theoretically produce in one season 12-1( 3 x 6 ) = 30 descendants; two litters of six young of equal sex ratio, with the three early-born young females each breeding later in the summer. The actual rate of increase achieved in a given year depends of course on the interaction of productivity and mortality, but the general consequences of this breeding strategy on a game estate may be illustrated as follows. In the present samples, between June and August the newly independent young females had several times the numbers and only half the mortality of the adult females (Tab 4). From Tab. 6 and Deanesly (1944) we may assume that both young and adult females are able to breed late in the season when conditions are favourable. Hence, the potential breeding stock in late summer contains many more females aged 3—4 months than 15 HOLARCTIC ECOLOGY 3:3 (1980) or more months of age, and it follows that, unless there is a gross increase in litter size with age, the greatest number of weasels comprising a peak population must have been produced by the early-born young females. The season of high mortality imposed by gamekeepers has largely passed by June (Fig. 1) and in any case affects mostly males; thus, the improved survival and precocity of the young females in summer ensures that the weasels killed in the spring trapping campaign can be more than replaced by autumn, certainly in years when rodents are abundant, and perhaps in most years. These preliminary results suggest that the traditional practice of weasel control on English game estates is probably effective only in the short term, if at all. Acknowledgements — This work was done at the Animal Ecology Research Group, University of Oxford, under the supervision of Dr H. N. Southern. Sources of material and ftnancial assistance are acknowledged in previous papers describing other aspects of the same colieclion (King 1975a, b, 1977). The 313 weasels from Sussex collected 1970-72, and the 78 from Aberdeenshire, were loaned by Dr P. J. Moors. This additional material has slightly altered the earlier estimates of mean age and mortality rates quoted earlier by King (1975a) and Corbet and Southern (1977). I am grateful to T. H. Blank and to the game estates whose data are quoted in Tabs 5, 7 and App. 1, respectively, and to Drs B. M. Fitzgerald, J. E. C. Flux and M. R. Rudge, for helpful MS criticism. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Ecography Wiley

Population biology of the weasel Mustela nivalis on British game estates

King, Carolyn M.
Ecography , Volume 3 (3) – Jul 1, 1980
Download PDF
10 pages

Loading next page...
 
/lp/wiley/population-biology-of-the-weasel-mustela-nivalis-on-british-game-RW8Er3YN6p

References (41)

Publisher
Wiley
Copyright
Copyright © 1980 Wiley Subscription Services, Inc., A Wiley Company
ISSN
0906-7590
eISSN
1600-0587
DOI
10.1111/j.1600-0587.1980.tb00722.x
Publisher site
See Article on Publisher Site

Abstract

King, C. M. 1980. Population biology of the weasel Mustela nivalis on British game estates. - Hotarct. Ecol. 3: 160-168. The mean year-class ratio in 6 samples of weasels (total n = 477) ranged from 59% to 84% young, and mean age from 0.79 to 1.16 yr. There was no difference in age structure between samples from 5 game estates and 1 reserve. Mean annual mortality of weasels of both sexes and all ages was 75-90%. In 12 pregnant weasels the mean number of embryos, observed from April to June, was 5.6. There was a defmite anoestrous season in winter. Records of the number of weasels killed on game estates show that weasels and stoats reacted differently to the first outbreak of myxomatosis in England in 1954; and that weasel populations are capable of enormous sudden increases (usually associated with "mouse years"). These increases are probably due largely to summer litters produced by precocious juveniles, and also to some adults breeding a second time (both possibilities confirmed from ovarian histology on this material). Traditional gamekeeping practice may influence the seasonal pattern of mortality of weasels (highest in spring) but its effects are probably short-lived. C. M. King, 3 Waerenga Road, Eastbourne, New Zealand. 1. Introduction Published studies of the population biology of Mustela nivalis are few, mainly because the two essential requisites, an accurate method of age-determination and unbiased samples, are difficult to supply. The age of weasels is difficult to determine (King 1980b), and in Britain, though large samples may be collected from co-operative gamekeepers, they are usually biased in several different ways. The literature does not contain a single analysis of the dynamics of a wild population of Musiela nivalis, not even a first approximation, although the commercially valuable mustelids have received closer attention, especially in the USSR (see, for example, many of the papers on sable, marten and ermine translated in King, 1975c, 1980a). In Britain the population biology of weasels is of considerable interest to the game industry, though the lack of any population data hitherto has prevented rational assessment of the traditional methods Throughout this paper, the name "weasel" refers only to Mustela nivalis, and "stoat" to M. erminea. Accepted 3 September 1979 © HOLARCTIC ECOLOGY 0105-9327/80/030160-09 S 02.50/0 usually employed for "vermin control". The present study sketches a first outline of this important missing information. Published estimates of age ratios for Mustela nivalis are difficult to compare with present results because not all authors defined their terms and methods in sufficient detail. Since first-year weasels* comprise about threequarters of the population, the total number at any time will depend largely on the production and survival of young, and the proportion of young in the sample will depend on the season of collecting and the definition of the youngest age class(es). Lockie (1966), Hansson (1968), Fog (1969), Barbu (1968) and Stubbe (1969) all used different definitions of age. Comparing them all shows only that all authors except Barbu and Hansson found the young animals to be in the majority. The distribution and abundance of weasels is clearly related to the local population of small rodents (Rubina I960, Parovshchikov 1963, Heptner et al. 1967, Eriinge 1974, King 1975a), though there are few direct data on the influence of nutrition on productivity (Tapper 1979), and none on its effect on survival. Reproductive cycles and development have been described by HJII HOLARCTIC ECOLOGY 3:3 (1980) median birth date assumed to be 1 June, although the season of births, beginning in April, is variable in length according to environmental conditions. The few old animals were all given an arbitrary age of 30 months. The weasel is a short-lived animal, but the age structures are expressed in year-class ratios, for practical reasons. The method of age determination has to work in chronological categories, because young weasels vary 2. Material and methods greatly in the age at which they reach reproductive Five of the samples (455 weasels) were collected by maturity, according to when in the season they were gamekeepers on estates in England and Scotland (num- horn. The year classes can of course be subdivided by bers 1, 2, 5, 6 and 7 on the map in King 1977). All were month, but if this is done, as in Tab. 4, the season at killed in Fenn traps, described by King and Edgar which the sample was collected determines the relative (1977), and were from populations regularly cropped prominence of each monthly or quarterly class. A samby trapping. A small sample of 22 weasels from ple collected in summer (June-August) contains many Wytham Woods, Berkshire, a reserve with no history of juveniles and few sub-adults, and vice versa one colweasel-trapping before 1968, is considered separately: lected in spring, though the same cohort of young, at these animals were either caught in box traps or shot. different stages in their development, is represented in Weasels which had been held in captivity or picked up both. By comparing samples in year-classes, differences due to season of sampling are reduced. Further, as dead were excluded. Most carcases were stored frozen. Autopsy proce- pointed out by Caughley (1977) and illustrated pardures included recording of measurements, food habits ticularly well in weasels, differences between age classes and Skrjabingylus nasicola (King 1977), and counting of spanning less than a year will relate more to season than embryos, and, for the first 171 specimens (samples 4, 5 to chronological age, so classes defined in years are not and 6, plus the first 55 of sample 1) also fatness, moult arbitrary; they are the only natural units of age. pattern, weight of gonads and histology of ovaries. PlaYoung weasels cannot be trapped before four weeks cental scars were not visible (Heidt 1970). Skulls were of age, when they first leave the nest, and are unlikely to cleaned by dermestid beetles. Demographic procedures enter traps before eight weeks, when they can kill for followed, where possible, the advice of Caughley themselves (East and Lockie 1964), Hence, all figures (1977). derived from trapped samples refer to mortality from at As on most game estates, the keepers concentrated least two months, not from birth. The samples are taken on trapping predators in spring, and hence the monthly to represent the standing age distribution in the living distribution of the samples is very uneven: 46% of the population from independence (at about three months) weasels examined were caught during March, April and onwards. May. This sampling programme was not ideal for The problems of population biology can be appopulation analysis, but the resulting material can at proached at several levels of detail, depending on the least be regarded as typical of the annual weasel harvest precision of the answer required and the difficulty of on many British game estates. extracting it (Caughley 1977). In the first instance, the All the game estates were mixed arable farmland at simplest approach is to treat all individuals of a given less than 500 m altitude: the largest sample came from sex or age group as if they were identical and to assume North Farm, Sussex, the study area described by Potts that the population statistics are representative and apand Vickerman (1974). Wytham is a largely deciduous ply as an average over several years. This approach is woodland described in detail by Elton (1966). There the most appropriate one for a pilot study, and is the were no estimates of weasel density on any of the game one used here. estates, but the resident population on a small part of Wytham was observed by livetrapping during 1968-70 (King 1975a). 3. Results The method of age-determination used is based on 3.1. Sex ratio the progression of changes in the shape of the skull established from specimens of known age (King 1977, The great excess of males in the samples (Tab. 1) was 1980b) Skulls were grouped first according to the analysed previously by King (1975b), who considered it month killed, and then into year-classes, divided at 1 more likely to be a sampling error arising from the difJune. Weasels in their first year are called "Young", ferences in the biology of males and females than a and those in their second year "Adults". Adults past measure of the real sex ratio of the population. Howtwo years of age ("Old") may be distinguished by wear ever, there were no seasonal variations in the relative of the carnassial teeth, which is not significant before number of males and females caught (Tab. 2); there was that age; but as this distinction is not precise it is used in no significant difference between males and females in only Tabs 1, 3 and 4. Mean age was estimated from a age structure as estimated from these trapped samples HOLARCTIC ECOLOGY 3;3 (t980) (1939), Deanesley (1944), and Heidt (1970); and variations in fur, bounty or vermin records by Parovshchikov (1963), Jefferies and Pendlebury (1968), and Hewson (1972). Tab. 1. Age distributions of weasels from 5 game estates and 1 reserve. Sussex Oxon. Game estates Northumb, Wigtown. Aberdeen Estates Combined Reserve Wytham Males Young Adult Old Total 164 39 39 30 Difference in age structure between estates (males): X = 2.75, 4 df (adult and old year classes combined), NS. ^ Mean age 0.86 yr Females Young Adult Old Total 66 13 5 84 «7 23 6 Difference in age structure between estates (females): X' = 4.82, 4 df, NS. Mean age 0.93 yr Total 248 42 54 33 78 455 22 Difference in age structure between sexes on the estates: X = 1.05, 1 df, NS. ^ Difference in age structure between estates and reserve (sexes pooled) X * 85.9 0.79 83.3 % males Mean age (yr) %young Ha/kill trap Collected Sample number (King 1977) 5,3 1968-72 1 ? 1968-70 2 66,7 8.1 1968-70 5 9.7 1969-70 6 74.5 0.88 78.5 1968-72 1970-72 1968-70 4 (Tab. 1); and sex ratio did not change with age (King 1975b). For the purposes of this paper I assume that both males and females have been sampled randomly, but at different levels, and so the results for the two sexes are presented separately below. 3.2. Age structure In every sample most of the weasels examined were in the first yearclass (Tab. 1). The actual ages of the old (2 years+) animals is unknown, but the age structures suggest that very few of them were much past their third year. The differences in age structure within the game estates, and between the game estates together and the reserve, were not significant (Tab. 1). The individual weasels examined lived on average about one year or less (the range of mean ages observed was 0.79-1,16 yr, mean 0.88 yr, in different samples, and 0.86 in males, 0.93 in females). However, the mean for the total population (including the pre-independence age classes, which was not sampled) was no doubt much less. The mean expectation of life at birth cannot be estimated if mortality before independence is un162 known. Weasels in captivity may live up to ten years (Frank, pers. comm.). The youngest weasels collected by gamekeepers were all over fifty days old, as all had completed their permanent dentition (East and Lockie 1964). This is probably not because younger weasels cannot be caught, but because the gamekeepers on the estates studied paid less attention to trapping in summer. In Wytham, one young male with deciduous canines still in place was caught in a live-trap in September 1969. A change in the proportions of the year-classes with time may indicate differential vulnerability to trapping (for example, in sable, many more young are caught at the beginning of the season than at the end: Mel'nikov 1975. In the present material, the proportion of young, in all samples combined, fluctuated irregularly through the year, with no obvious decline with time since the season of births (Tab. 2). The mean annual mortality rates of the two year classes were similar, in both males and females (Tab. 3). If regular trapping were having any effect upon the target population, e.g. by stimulating breeding or immigration, or reducing the mean lifespan, a higher HOLARCnC ECOLOGY 3:3 (1980) Tab, 2. Proportion of young and adult weasels and sex ratios through the year (Game estates combined). Age Tab. 4. Seasonal pattern of mortality in weasels on game estates. Capture date Males First year Jun-Aug (Summer) Sep-Nov (Autumn) Dec-Feb (Winter) Mar-May (Spring) Second year Jun-Aug Sep-Nov Dec-Feb Mar-May Older (all seasons) Females First year Jun-Aug Sep-Nov Dec-Feb Mar-May f. 339 293 245 193 Young Jun Jul Aug Sep Ott Nov Dec Jan Feb Mar Apr May 15 32 18 II 18 32 34 24 18 82 54 36 Total %Young 29 43 23 U 22 36 36 32 21 100 69 45 Sex (from King 1975b) n CfCf Total %cfcf U 1.00 0.86 0.72 0.57 d, 0.14 0.14 0.15 0.37 q. 0.14 0.16 0.21 0.65 51.7 74.4 78.3 100 81,8 88,9 94.4 75,0 85.7 82.0 78,3 80,0 76.5 58,8 70.0 100 77,8 77.2 73.7 51.5 80,0 78,0 74,7 73.7 0.20 0,05 0. 15 0.02 0. 13 0.02 0. 11 0,09 0.02 0,82 trapping pressure (expressed as ha/trap) should be Jissociated with a lower mean age. The trapping pressure was estimated for only three of the game estates, and, within the range observed, was not obviously related to variations between estates in mean age (Tab. 1). The mean age in the Wytham Reserve was 1.02 yr (n = 22), which was within the range of mean ages observed on the game estates. The significance of the differences between these figures was not tested because the actual ages were not known: the range of birth-dates (4—5 months) is about the same as the range of differences observed. 3.3. Seasonal (three-monthly) variations in mortality rale (qj 0.15 O.Il 0.22 0.58 0.32 0 0.24 0.62 Second year Jun-Aug Sep-Nov Dec-Feb Mar-May Older (all seasons) mathematically related, but q^ is the only one plotted, as it is the statistic least affected by errors of age-determination and sampling bias (Caughley 1977), 3.4. Natural mortality The mortality pattern of the weasels on the game estates showed a well-defined peak in the spring of each year (Fig. I). However, although this suggests that the activities of gamekeepers may affect the seasonal distribution of mortality, it does not imply that they necessarily affect the annual level of mortality. The q, curve for most mammals is U-shaped; the initial drop is absent from Fig, 1 because q» data derived from trapped samples take no account of mortality before independence. All the colums of Tab. 4 are There is little evidence from this study of the causes of natural mortality in weasels. Of eight weasels found in the pellets of tawny owls in Wytham (Southern, pers. comm.), five were female, two male and one undertermined; two were adult, two young and four undetermined. The nematode parasite Skrjabingylus nasicola causes considerable damage to the sktills of weasels, and 77-100% of skulls in the present samples were infested, but there was no evidence that infested weasels were smaller, lighter, leaner or died younger than uninfested ones (King 1977). Tab. 3. General pattern of mean annual mortality rates of male and female weasels on game estates. Age (yr) Frequency fx Males 0.25-1 1-2 339 69 8 116 29 6 Survival I. 1.00 0.20 0.02 1.00 0.25 0.05 Mortality d. 0,80 0.18 0.80 0,90 Mortality rate Survival rate P« 0,20 0.10 2+ Females 0.25-1 1-2 2+ HOLARCTIC ECOt-OGY 3:3 (1980) Tab. 5. Records of weasel pregnancies and lactations. Pregnancies Date 19 Apr 1953 25 Apr 1958 26 Apr 1971 27 Apr 1972 during april 1969 1 May 1969 10 May 1969 12 May 1969 17 May 1962 21 May 1972 during May 1969 26 Jun 1958 Sample area Hampshire' Hampshire Sussex Aberdeen Sussex Northumberland Northumberland Wytham Hampshire Sussex Sussex Hampshire Body wt (g) No, of embryos 0-2 ADULTS Mean Lactations Sample area Sussex Dates Fig. 1. Seasonal variation in mortality rate, q^, in male (•) and female (o) weasels on game estates. Sm: summer. A: autumn, W: winter, Sp: spring. 3.5. Reproduction 3.5.1. Litter size The mean number of embryos was 5.6 (Tab. 5). As found by Deanesly (1944) the corpora lutea in weasels persist for some lime after parturition (in one pair of ovaries I sectioned, corpora lutea were still present in October, and in another pair, two generations of corpora lutea were found together in July); but the poor quality of the histological material made it impossible to count them. 3.5.2. Breeding season Embryos were observed between April and Jtjne, and lactation between April and July with one (in the south of England) lactating in October (Tab. 5). The end of the season is difficult to determine, because few weasels were collected from mid-summer onwards, and in any case the length of the season probably varies from year to year. Tapper (1976) made a special effort to collect weasels in summer from North Farm, Sussex, and some of those he caught are included in this material. In July and August of 1971 and 1972, 19 females were collected, of which 5 were rotten. As the ovaries were not sectioned, the breeding condition of the remaining 14 could not be fully determined. From macroscopic signs, 2 of the 14 were lactating, and one was in oestrus, but none was visibly pregnant. At Wytham, the longest resident adult female, 9 18, was pregnant when livetrapped on 26 August 1969, but had apparently lost the litter by the time she was next captured three weeks later (King 1975a). The non-fertile season was well marked. In 26 females collected between November and March inclusive, whose ovaries were sectioned, all but one were anoestrous (Tab. 6). The annual cycle of testis weight followed closely the description of Hill (1939). 5 Apr 1971 , 7 May 1971, 1 Jun 1971, 23 Jun 1972, 30 Jun 1971, 19 Jul 1971,24 Jul 1972,14 Oct 1971 Northumberland 9 Jun 1969 Hampshire 2 Jun 1949 1, Hampshire records from Blank (unpubl.). 3.5.3. Fertilily and age Tab. 6 shows the relationship between breeding condition, season and age in the 45 females whose ovaries were sectioned. All three young females caught before the end of the breeding season were fertile, though none had visible embryos. In 16 mature females caught after April, all except one were in breeding cotidition, and the only one caught in July was already fertilised for her second pregnancy; only five of these had visible embryos. Tab. 6. Reproductive condition and age cf 45 female weasels whose ovaries were sectioned. A, Young 9 9 caught before 31 Oct of [heir first year: B, Females of all ages (mature and immature) caught 1 Nov~31 Mar inclusive: C, Females of all ages (all mature) caught after 1 Apr: Fertile (in oestrus or with corpora lutea) 3, Infertile 0. Anoestrous 25 Oesirous 1 (12 Mar) Anoestrous 1 (13 Apr) Oestrous 5 (16 Apr to 9 May) Between ovulation and lactation of first pregnancy of season 9 (28 Apr-9 Jun) Between ovulation and lactation of second pregnancy of season 1 (7 Jul) Infertile 0 Total 16 HOLARCTIC ECOLOGY 3:3 {19B0) Tab. 7. Recent variations in the number of weasels killed per year on three of the game estates studied. Year Sussex Northumberland Wigtownshire (14000 ha) (1200 ha) (4000 ha) none need be expected in these samples, for three reasons. Trapping need have no effect on age ratios if all ages are equally liable to be caught (Caughiey 1974), or if the artificial mortality imposed by gamekeepers is insignificant; and in theoretical populations with similar patterns of age-dependent survival but different rates of increase, the age-ratios can be unrecognisably different (Caughley 1977). Hence, age-ratios cannot be interpreted without a knowledge of r. 4.2. Mortality and survival 3.5.4. Uterus weights Uteri were weighed complete with ovaries attached. The mean weight of 26 anoestrous uteri was 43 mg, range 24—76 mg: of 13 uteri, determined from ovarian sectioning and the appearance of the vulva to be pre- or post-oestrous, 137 mg, range 45-275 mg: of 5 uteri containing embryos, 2.17-6.68 g. 3.5.5. Population Fluctuations The densities of weasels on the five game estates were not known, but they were certainly not stable from year to year. Records of the annual kill of weasels by gamekeepers, which are Ihe crudest form of population estimate, were available from three of the estates for the years preceding the study (Tab. 7). They show that the "harvest" of weasels is unstable, frequently varying between years by a factor of 2 or 3. This instability means that data on mortality and reproduction based on specimens collected over several years provide only a general picture. Tapper (1979) describes the responses of a local weasel population lo changes in its food supply. 4. Discussion 4.1. Age slnidure The pronounced seasonal variation in mortality rate of weasels on game estates is probably a largely artificial consequence of the extra trapping effort traditionally made by gamekeepers in spring. However, weasels also change their behaviour with the onset of the breeding season in spring, and this may increase their "trappability" and also, on untrapped areas, increase the risk of death from factors other than traps. Several resident male weasels in the reserve died after drastic loss of weight at this season (King 1975a). In the combined sample from the game estates, the survival of female weasels from independence to one year old was slightly higher (0.25) than that of males (0.20: Tab. 3), though the difference in the age distributions of the sexes was not statistically significant (Tab. 1). Gamekeepers invariably kill about three times more male weasels than females (King 1975b) which means that more females than males escape the trapping season and later die from natural causes. The small difference in these survival figures perhaps reflects the lower proportion of the female population vulnerable to the risk of being trapped. In contrast, live-trapping studies of undisturbed weasel populations have frequently recorded periods when females were apparently scarce or completely absent (Lockie 1966, Eriinge 1974, King 1975a, Moors pers. comm.). The difference could perhaps be due to the beneficial effects of cropping. Cropped populations often show improved condition and vigour; the effects of traditional weasel trapping, highly selective for males (Tab. 2) would benefit females the most, particularly as in weasels the nonbreeding females are socially subordinate to males and inferior to them in hunting ability (Eriinge 1975), 4.3. Reproduction Stubbe (1969) commented that the high proportion of young in samples of weasels implies high productivity in weasel populations; but immigration stimulated by constant cropping and selective trapping of young can result in an observed proportion of young far exceeding the possible productivity of a population. Hence, the productivity of the population sampled can be estimated only from fecundity data obtained from examining the females. In the present material, only 145 of the total of 535 weasels collected were females, of which only 8 The age structures observed correspond reasonably well with what would be expected in a small mammal capable of rapid reproduction, and with field observations. Few resident weasels hold their territories for more than one year (Eriinge 1974, King 1975a), and none has been known to live in the wild for more than three years (Lockie 1966). Because of the variable occurrence of summer litters, a theoretical age pyramid cannot be constructed for weasels to compare with that calculated for stoats by van Soest and van Bree (1970). There was no difference between the age-structures of the weasels from the estates and the reserve, but HOLARCTIC ECOLOGY 3:3 (1980) were pregnant. Even when sampling is carefully controlled, female weasels are apparently particularly difficult to catch when pregnant (Deanesly 1944, Behnke 1966). The only recorded count of corpora lutea in British weasels is that of Deanesly (1944), who found an average of 7.1 corpora lutea in 32 pregnant weasels examined. The mean number of embryos found in this study {5.6} was comparable to the 6.4 found by Deanesly. The mean number of young born, In 17 litters recorded in Britain, was 6.2 (range 4-8): if a further 22 litters from other parts of the species distribution are included, the mean is 5.0 (King 1975b). The limits of the anoestrous season in British weasels seem to be fairly definite (Deanesly 1944, Tab. 5, 6), in contrast to European nivalis (Stubbe 1969, Heptner et al, 1967) and American rixosa (= nivalis) (Hall 1951). Adult female weasels are known to be capable of producing two litters a year (Heidt 1970), and young females bom early in the season may themselves reproduce later in the summer (Deanesly 1944). Such extended breeding seasons are associated with years when small rodents are very abundant (Hensel 1881) and so also are records of very large litters, which may be up to at least fifteen young (Heptner et al. 1967, Eitzgerald in press). Second litters are possible in "lemming years" even in the short summers of the far north, as Frank (1974) reports that a captive female of the small boreal form M. nivalis rixosa regularly came into post-partum oestrus before the end of lactation, and could rear two litters in 5.5 months. However, monthly trapping records, showing two peaks in capture rate in April and July, do not infer that second litters are normal in Britain (Jefferies and Pendlebury 1968). The first peak is too early to represent dispersing young of the first litter (Tabs 5, 6) and may be more related to spring changes in behaviour and trappability mentioned above. In poor years, many adult weasels may fail to breed altogether (Tapper 1979). 4.4. Population fluclualions Appendix I, Vermin records from Elveden estate, Norfolk (approx 9300 ha). Year Number of animals killed No keepers M. anc warreners M. nivalis erminea rabbits employed 83 92 88 73 88 56 59 168 56 73 75 85 90 34 7 10 11 10 4 3 14 8 20 17 18 I t 17 35 22 42 81 37 16 26 64 22 23 24 715 766 859 925 1013 825 459 419 422 439 596 655 643 434 409 438 569 517 495 855 624 568 735 647 603 645 727 825 257 146 98 80 62 71 40 58 75 170 251 257 166 138 164 225 270 168 249 348 318 448 51563 61571 55986 51777 57240 48602 58561 54169 52606 51163 48870 43416 42294 31706 20348 15765 13218 22744 40376 47742 50009 53793 47645 27381 23535 19528 19839 8601 401 1864 4968 5016 3037 9001 3962 5105 4864 2607 680 849 949 1147 1760 1354 2954 3158 4083 8917 14378 15698 35 Records not available 37 37 33 13 13 14 14 16 26 29 27 27 22 26 25 26 26 19 19 t9 20 19 23 23 22 19 15 11 Keepers on War Service First outbreak of myxomatosis U2 The annual variation in the number of weasels killed on British game estates illustrated in Tab, 7 is comparable with figures published elsewhere: for example, for the ten years 1947-48 to 1956-57 inclusive. Anon (1960) reports the annual kill of weasels on 1336 ha in Hampshire as 82, 106, 185, 96, 84, 110, 46, 64, 53 and 68 respectively. Series of figures showing a similar range of variation can be found in the vermin books of almost any estate, and well back into last century (Middleton 1934). Ocassionally, weasels are capable of sudden quite startling irruptions. The next figure in the above series, for 1957-58, is 348, a five-fold increase over that of the previous year. This is almost certainly due to the consequences of myxomatosis, which were demonstrated in records from all over the country (Jefferies and 166 Pendlebury 1968, Hewson 1972, Appendix 1.) As the rabbits declined, followed by the stoats, the herb layer of the vegetation became luxuriant, and populations of small rodents reached peak levels (Fenner and Ratciiffe 1965, Hewson and Kolb 1973), followed by (he weasels. Over much of the country the main result was a reversal of the previous ratio of stoats to weasels (Craster 1970, Hewson 1972). The increase in weasels was probably due mainly to the improvement in their HOt_ARCTIC ECOLOGY 3:3 (1980) food supplies, though the removal of interspecific competition from stoats may have helped (King and Moors 1979). Appendix 1 suggests that, at least over the last twenty years, the harvest of weasels on this estate has been higher than usual every three or four years. The greatest value of vermin records is that they are an eloquent warning against any facile assumption of a stationary population in short-term or localised samples of weasels. Fur and bounty records in general are often unreliable, because the incentive to set traps changes with the density of the target population; but the figures provided by gamekeepers are probably less subject to this error, since setting traps is done in the course of other work {Hewson and Kolb 1973). It seems safe to assume that weasel populations are actually at least as unstable as these records suggest, and that the rate of increase is usually either positive or negative, but very seldom zero. This means that lifetables more accurate than the preliminary ones presented here can be constructed only by studying marked animals (Caughley 1977). Sudden large population increases in weasels are often associated with years when small rodents are abundant (Swanson and Fryklund 1935, Lokemoen and Higgins 1972, McLean et al. 1974, Goszczynski 1977). This could readily be understood if the effect of extra food is merely to decrease the high mean annual mortality rate; but in weasels, improved survival can be augmented not only by an increase in fecundity, which occurs in other carnivores and may reasonably be expected, but also the capacity (unique among mustelids) of weasels to produce extra litters in good seasons. 4.5. Implicalions for the preservation of game The flexible reproductive strategy of the weasel allows it to avoid surplus production in poor years, and to exploit favourable circumstances very quickly, despite the attentions of gamekeepers. This latent breeding capacity is certainly enough to account for large population peakssuchas that of 1957-58 reported by Anon (1960) and others. In ideal conditions in Britain a single female weasel could theoretically produce in one season 12-1( 3 x 6 ) = 30 descendants; two litters of six young of equal sex ratio, with the three early-born young females each breeding later in the summer. The actual rate of increase achieved in a given year depends of course on the interaction of productivity and mortality, but the general consequences of this breeding strategy on a game estate may be illustrated as follows. In the present samples, between June and August the newly independent young females had several times the numbers and only half the mortality of the adult females (Tab 4). From Tab. 6 and Deanesly (1944) we may assume that both young and adult females are able to breed late in the season when conditions are favourable. Hence, the potential breeding stock in late summer contains many more females aged 3—4 months than 15 HOLARCTIC ECOLOGY 3:3 (1980) or more months of age, and it follows that, unless there is a gross increase in litter size with age, the greatest number of weasels comprising a peak population must have been produced by the early-born young females. The season of high mortality imposed by gamekeepers has largely passed by June (Fig. 1) and in any case affects mostly males; thus, the improved survival and precocity of the young females in summer ensures that the weasels killed in the spring trapping campaign can be more than replaced by autumn, certainly in years when rodents are abundant, and perhaps in most years. These preliminary results suggest that the traditional practice of weasel control on English game estates is probably effective only in the short term, if at all. Acknowledgements — This work was done at the Animal Ecology Research Group, University of Oxford, under the supervision of Dr H. N. Southern. Sources of material and ftnancial assistance are acknowledged in previous papers describing other aspects of the same colieclion (King 1975a, b, 1977). The 313 weasels from Sussex collected 1970-72, and the 78 from Aberdeenshire, were loaned by Dr P. J. Moors. This additional material has slightly altered the earlier estimates of mean age and mortality rates quoted earlier by King (1975a) and Corbet and Southern (1977). I am grateful to T. H. Blank and to the game estates whose data are quoted in Tabs 5, 7 and App. 1, respectively, and to Drs B. M. Fitzgerald, J. E. C. Flux and M. R. Rudge, for helpful MS criticism.

Journal

Ecography – Wiley

Published: Jul 1, 1980

There are no references for this article.